Clinical Implications of the New 2010 WHO Reference Ranges for Human Semen Characteristics

Note: Postings to Androlog have been lightly edited before publication. Although much emphasis in society is placed on female infertility, male factor issues play an etiologic role in a substantial portion of the 10% to 15% of couples who have trouble conceiving (Singh and Jaiswal, 2011). The root of impaired male reproductive health is often difficult to specify and can sometimes be multifactorial, with environmental, nutritional, genetic, and behavioral issues contributing to the overall problem (Hwang et al, 2011). The challenge for urologists is often this: providing couples with an accurate assessment of the male partner's reproductive potential. To this end, urologists and other clinicians often rely upon published reference ranges for human semen characteristics in providing individual males prognostic information about their fertility. In 2010, the World Health Organization (WHO) published reference values for semen, and these have been widely disseminated for clinical use. However, with the arrival of these new reference ranges come significant concerns about their clinical relevance and applicability. Throughout the years, the clinical evaluation of the infertile male has come to include many costly and complicated tests, including hormonal assays and genetic analyses. However, the most important laboratory test is the semen analysis. In 1951, Dr John MacLeod published in Fertility and Sterility objective measurements of semen after analysis of 1000 fertile men and 800 infertile men. For each semen sample, he assessed ejaculate volume, sperm count, sperm motility, and sperm morphology as potential factors contributing to infertility. In his conclusions, MacLeod found that men in the count range between 1 and 20 million sperm/mL were 3 times more likely to be infertile than fertile. Yet a significant portion of the men in that category were considered fertile, underscoring the lack of specificity for low sperm concentration to serve as a clear marker of male factor infertility. Over the years, the WHO used thresholds similar to those in MacLeod's original paper to define semen parameter reference ranges. In 2010, the WHO published a new set of reference values for semen analyses based on the assessment of 4500 fertile men from 14 countries (Cooper et al, 2010). These reference ranges are also published in the widely utilized WHO Laboratory Manual for the Examination and Processing of Human Semen, 5th edition (WHO, 2010). In this edition of the WHO manual for semen analysis, Table A1.2 shows values for semen parameters of volume, total sperm number, sperm concentration, total motility, progressive motility, nonprogressive motility, immotile spermatozoa, vitality, and normal forms for a cohort of men whose partners became pregnant within 12 months of discontinuing contraceptive use. In addition, Table A1.1 states the lower reference limits, the 5th centiles, and 95% confidence intervals for semen characteristics. The assignment of these 5th-centile values as “lower reference limits” was new to this edition of the WHO manual and is the source of much controversy. Although the 5th centile is often used as a quantitative value for statistical analysis throughout science, its assignment as a discriminating cutoff value in male reproductive potential is novel. More specifically, the 5th-centile values for semen parameters were generated on the basis of broader statistical norms and not on the basis of any clinical outcomes from fertile and infertile men. In other words, there is no clear evidence that application of these values effectively segregates men on the basis of their fertility, yet that is exactly how these new ranges are being applied clinically throughout the world. As noted by Niederberger (2011), although 5th-centile values are commonly used as cutoff markers in statistics, the ability of this arbitrarily assigned cut point to provide meaningful information about a male's fertility potential is questionable. Soon after publication of the 5th edition of WHO manual for semen analysis, discussion entries on the new guidelines were posted on Androlog. The following excerpts from the Androlog website summarize the discussion by colleagues on the interpretation of the new guidelines and its implications on clinical practice. Dr Alayman Hussein started the discussion with the following posting: We are seeing and treating many patients having an infertility problem according to the previous edition of WHO manual. Now, according to the WHO manual 5th edition, they are considered normal. My question is how to convince these men that they are normal? In response to Dr Hussein's question, Dr Craig Niederberger replied: It is wrong to equate WHO threshold values with normal ones. In the 5th edition of its manual, the WHO chose to emphasize the 5th centile as threshold values (Table A1.1, page 224). Normal values would be closer to the 50th centile (Table A1.2, page 225). For example, the 5th centile for sperm density is observed to be 15 million/mL, whereas the 50th centile value is 73 million/mL. There is an enormous space between the 5th and 50th centiles where many infertile men may lie, as evidenced by the substantial overlap in the frequency histograms between fertile and infertile men according to semen analysis parameters going back to MacLeod's reports over half a century ago. One good strategy to resolve infertility by semen analysis parameters is to apply Guzick and coinvestigators' approach of setting two thresholds, one below which a man is likely to be infertile, and above another where he is likely to be fertile. (Guzick and colleagues' approach was to use classification and regression tree, or CART, analysis.) The publication Dr Niederberger cites is the 2001 article by Guzick et al titled “Sperm Morphology, Motility, and Concentration in Fertile and Infertile Men” in the New England Journal of Medicine. In this article, Guzick et al analyzed 2 semen samples from male partners in 765 infertile couples and 696 fertile couples. Guzick and colleagues showed that in terms of sperm concentration, men with sperm concentration over 48 million/mL were likely fertile whereas men below 13.5 million/mL were likely infertile. Men whose sperm concentrations were in between the 2 values represent an indeterminate range. The authors noted that although low sperm concentration values increased the chance that a male factor contributes to a couples' infertility, there was considerable overlap between the frequency distributions of fertile men and infertile men. Further, low values for sperm concentration, motility, or morphology that fall in the subfertile range are not absolutes for infertility because of the large frequency overlap in male populations. It remains to be seen just how many institutions will adopt the new reference values. More importantly, how these reference values will be utilized by health care providers in clinical decision making is still unknown. Many clinicians are concerned that the new reference values will shift the burden of treatment to the female partner, with more men being considered “normal” as their semen parameters fall into the >5th centile range. Specialists should thus proceed with caution in using these new reference ranges when providing care for infertile males. These cutoff values should be regarded as statistical markers from a large population of fertile men, not proven boundaries between “fertile” and “infertile.”